INDEX

Explanations

terms related to feeling impressed or impressing others

oai_token-act-pair · gpt-3.5-turbo

expressions of admiration or being impressed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ignty

-0.69

space

-0.69

adobe

-0.67

 sovere

-0.65

 landfill

-0.58

 hijacked

-0.58

practice

-0.57

ichick

-0.57

iny

-0.57

 space

-0.56

POSITIVE LOGITS

ments

1.00

ively

0.99

mented

0.99

 upon

0.88

ingly

0.87

 enough

0.85

Reviewer

0.83

ment

0.83

mentation

0.82

 nonetheless

0.77

Activations Density 0.057%

terms related to feeling impressed or impressing others

expressions of admiration or being impressed

No Comments

No Known Activations