INDEX

Explanations

phrases related to defining, describing, or knowing something or someone

oai_token-act-pair · gpt-3.5-turbo

phrases indicating inclusivity or acceptance in social 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 corrid

-0.69

ards

-0.62

©¶æ

-0.61

nesday

-0.60

 grape

-0.59

aband

-0.58

 laun

-0.58

oes

-0.57

ordes

-0.56

POSITIVE LOGITS

liest

0.87

acid

0.78

wikipedia

0.68

atown

0.67

velength

0.65

except

0.61

Na

0.61

pected

0.60

avering

0.60

ĸļ

0.59

Activations Density 0.320%

phrases related to defining, describing, or knowing something or someone

phrases indicating inclusivity or acceptance in social contexts

No Comments

No Known Activations