INDEX

Explanations

terms related to statistics and numerical comparisons

oai_token-act-pair · gpt-3.5-turbo

phrases related to belonging or association with various groups and entities

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

 sidx

-0.82

helm

-0.81

ãĤ©

-0.81

ãĤ³

-0.76

wcsstore

-0.76

ãĤ¤ãĥĪ

-0.75

nil

-0.74

HCR

-0.74

ãĤ£

-0.70

strong

-0.70

POSITIVE LOGITS

 affected

0.94

 Rate

0.68

 stricken

0.68

 Individuals

0.67

 Fiorina

0.65

 Rates

0.65

 Kern

0.64

 impacted

0.63

 Fairfax

0.63

 Americans

0.63

Activations Density 0.735%

terms related to statistics and numerical comparisons

phrases related to belonging or association with various groups and entities

No Comments

No Known Activations