INDEX

Explanations

important events or decisions

oai_token-act-pair · gpt-3.5-turbo

significant announcements or changes in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Pros

-0.68

-0.60

training

-0.60

gom

-0.58

Narr

-0.58

âĹ¼

-0.57

layer

-0.56

 dying

-0.56

 invincible

-0.55

ocrates

-0.54

POSITIVE LOGITS

 coincides

1.36

 coincided

1.31

 underscores

1.24

 signifies

1.20

 comes

1.17

 reflects

1.15

 reinforces

1.13

 brings

1.13

 represents

1.11

 prompted

1.10

Activations Density 0.238%

important events or decisions

significant announcements or changes in context

No Comments

No Known Activations