INDEX

Explanations

information related to individuals, their actions and associations, especially in a legal or political context

oai_token-act-pair · gpt-3.5-turbo

entities and individuals involved in criminal or suspicious activities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tnc

-0.77

 drought

-0.70

peat

-0.69

 reflection

-0.69

 FINAL

-0.68

 celebration

-0.68

mma

-0.67

umph

-0.66

 mourning

-0.66

 applause

-0.65

POSITIVE LOGITS

 allegedly

1.44

 testified

1.30

 oversaw

1.26

 laund

1.22

 pled

1.21

 alleges

1.20

 reportedly

1.19

 pleaded

1.17

 funnel

1.15

aunder

1.13

Activations Density 0.504%

information related to individuals, their actions and associations, especially in a legal or political context

entities and individuals involved in criminal or suspicious activities

No Comments

No Known Activations