INDEX

Explanations

verbs related to actions or commands

oai_token-act-pair · gpt-3.5-turbo

verbs indicating actions or activities

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

Applic

-0.93

HCR

-0.79

FTWARE

-0.75

amation

-0.65

 Apart

-0.64

CLR

-0.63

WER

-0.63

immer

-0.62

 guiActiveUnfocused

-0.60

 Calculator

-0.60

POSITIVE LOGITS

backs

0.75

 ourselves

0.71

ings

0.67

fully

0.66

iens

0.64

idas

0.64

 yourself

0.64

gha

0.63

wright

0.62

birds

0.61

Activations Density 0.354%

verbs related to actions or commands

verbs indicating actions or activities

No Comments

No Known Activations