INDEX

Explanations

dates in the format "YYYY-MM-DD" with high emphasis on the year

oai_token-act-pair · gpt-3.5-turbo

negative sentiments or adverse remarks

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 ACTIONS

-0.68

ç¥ŀ

-0.65

iosis

-0.63

 birthplace

-0.62

 Opportun

-0.61

esan

-0.61

 grop

-0.60

;;;;;;;;;;;;

-0.59

 someday

-0.58

ãĤ§

-0.58

POSITIVE LOGITS

1.17

1.16

1.13

1.12

1.10

1.08

1.07

Activations Density 0.053%

dates in the format "YYYY-MM-DD" with high emphasis on the year

negative sentiments or adverse remarks

No Comments

No Known Activations