INDEX

Explanations

phrases indicating varying degrees of comparison or importance

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize intensity or degree

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

ãĤ¼ãĤ¦ãĤ¹

-0.61

 Halls

-0.60

enhagen

-0.60

ãĥį

-0.59

KM

-0.58

 acre

-0.58

MORE

-0.56

REC

-0.55

 SHARES

-0.54

ãĥ«

-0.54

POSITIVE LOGITS

apy

1.01

othes

0.97

oths

0.97

bered

0.94

othe

0.87

zin

0.79

aker

0.76

iled

0.76

aps

0.76

oner

0.74

Activations Density 0.045%

phrases indicating varying degrees of comparison or importance

phrases that emphasize intensity or degree

No Comments

No Known Activations