INDEX

Explanations

words related to bureaucratic processes or procedures

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "red" 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 @ 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

 Lank

-0.80

ILA

-0.80

Ö¼

-0.78

UGH

-0.75

agall

-0.74

gerald

-0.74

awaru

-0.72

ernel

-0.71

=-=-

-0.71

POSITIVE LOGITS

rawn

1.15

oubt

1.13

neck

1.12

efined

1.06

iscovery

1.01

headed

0.99

iscovered

0.99

oub

0.98

iscover

0.98

 velvet

0.95

Activations Density 0.022%

words related to bureaucratic processes or procedures

occurrences of the word "red" in various contexts

No Comments

No Known Activations