INDEX

Explanations

phrases related to depth, either literal or metaphorical

oai_token-act-pair · gpt-3.5-turbo

references to varying depths in underwater or geological contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Faith

-0.86

å¥³

-0.86

advertising

-0.80

Bey

-0.73

neg

-0.73

Dash

-0.71

committee

-0.71

DEF

-0.70

ussy

-0.70

DOM

-0.69

POSITIVE LOGITS

 caves

1.12

 depths

0.98

 canyon

0.96

 cliffs

0.95

 lake

0.90

 altitude

0.89

 deepest

0.84

 lakes

0.84

 temperatures

0.84

 rocky

0.83

Activations Density 0.095%

phrases related to depth, either literal or metaphorical

references to varying depths in underwater or geological contexts

No Comments

No Known Activations