INDEX

Explanations

terms related to actions or decisions (e.g., reversal, developments, finding, decision)

oai_token-act-pair · gpt-3.5-turbo

key events or actions related to decisions and developments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Native

-0.74

Tang

-0.64

Zip

-0.63

avis

-0.63

;;

-0.61

burning

-0.61

Scotland

-0.61

Choose

-0.60

 Katrina

-0.60

ochet

-0.60

POSITIVE LOGITS

 coincides

0.84

 coincided

0.81

 prompted

0.77

 relates

0.76

 underscores

0.75

 embold

0.75

 includes

0.73

 consists

0.73

 amounted

0.73

 consisted

0.73

Activations Density 0.391%

terms related to actions or decisions (e.g., reversal, developments, finding, decision)

key events or actions related to decisions and developments

No Comments

No Known Activations