INDEX

Explanations

words with negative connotations or describing negative characteristics/actions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "bad" in various contexts

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

ĸļ

-0.91

raltar

-0.84

ensional

-0.79

ittees

-0.78

conservancy

-0.77

aukee

-0.77

eters

-0.76

olate

-0.76

ynthesis

-0.74

illation

-0.74

POSITIVE LOGITS

dest

1.08

dies

1.04

die

0.98

gered

0.92

 karma

0.91

ger

0.89

 luck

0.86

 manners

0.80

 publicity

0.78

ged

0.78

Activations Density 0.025%

words with negative connotations or describing negative characteristics/actions

instances of the word "bad" in various contexts

No Comments

No Known Activations