INDEX

Explanations

phrases or words related to encoding or decoding information

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "odes" or similar variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Jimmy

-0.71

 disclaimer

-0.64

 Geral

-0.64

 duck

-0.64

 lawy

-0.63

 Rita

-0.62

 drinking

-0.60

 dred

-0.60

 standing

-0.60

McG

-0.60

POSITIVE LOGITS

odes

4.82

ode

2.92

ODE

2.19

oded

1.92

oding

1.82

oder

1.58

od

1.22

odic

1.18

ods

1.15

otes

1.11

Activations Density 0.006%

phrases or words related to encoding or decoding information

occurrences of the term "odes" or similar variations

No Comments

No Known Activations