INDEX

Explanations

words related to negative characteristics or actions

oai_token-act-pair · gpt-3.5-turbo

instances of the prefix "un" indicating negation or lack 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

uyomi

-0.91

anwhile

-0.89

hyde

-0.84

å§«

-0.80

 Nanto

-0.79

ONES

-0.76

*/(

-0.74

 Opera

-0.71

OPLE

-0.69

MIC

-0.69

POSITIVE LOGITS

avour

1.08

athom

0.98

ortun

0.95

ruit

0.91

ishable

0.91

riend

0.91

ortunately

0.89

ancies

0.87

requ

0.86

aced

0.84

Activations Density 0.002%

words related to negative characteristics or actions

instances of the prefix "un" indicating negation or lack in various contexts

No Comments

No Known Activations