INDEX

Explanations

mentions of historical locations or figures

oai_token-act-pair · gpt-3.5-turbo

historical references and significant ancient locations

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

 rollout

-0.86

 ramps

-0.84

 NETWORK

-0.84

 dashboard

-0.78

 actionGroup

-0.78

 stickers

-0.77

 lasers

-0.77

 Walmart

-0.77

 networking

-0.77

 interns

-0.77

POSITIVE LOGITS

Ã»

1.31

Ã¦

1.20

anus

1.16

á¸

1.13

Ã¼

1.12

atha

1.12

ocrates

1.10

Ã¶n

1.09

olkien

1.08

Ã¢

1.07

Activations Density 0.505%

mentions of historical locations or figures

historical references and significant ancient locations

No Comments

No Known Activations