INDEX

Explanations

descriptions of observations or actions related to learning or understanding

oai_token-act-pair · gpt-3.5-turbo

concepts related to observation and consensus

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

stanbul

-0.73

)]

-0.73

]);

-0.69

berra

-0.68

)}

-0.68

 meanwhile

-0.68

?)

-0.67

Wem

-0.67

resy

-0.64

yss

-0.64

POSITIVE LOGITS

 crap

0.65

 incredibly

0.63

 pretty

0.63

 insanely

0.61

oft

0.61

 horrend

0.60

 EVERY

0.59

 absolutely

0.59

 horrendous

0.59

 shit

0.58

Activations Density 1.770%

descriptions of observations or actions related to learning or understanding

concepts related to observation and consensus

No Comments

No Known Activations