INDEX

Explanations

numerical sequences in the format of "number-number"

oai_token-act-pair · gpt-3.5-turbo

references to sports scores and outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 wedd

-0.71

auga

-0.66

 Stub

-0.66

 "$:/

-0.66

 Sidd

-0.65

 Strait

-0.63

 Lenn

-0.63

 admiration

-0.63

 Lucas

-0.63

sul

-0.62

POSITIVE LOGITS

degree

1.06

digit

0.96

inch

0.96

year

0.95

million

0.93

minute

0.93

Ele

0.93

tier

0.91

hour

0.91

ohm

0.90

Activations Density 0.051%

numerical sequences in the format of "number-number"

references to sports scores and outcomes

No Comments

No Known Activations