INDEX

Explanations

phrases related to observations and descriptions of things or situations

oai_token-act-pair · gpt-3.5-turbo

occurrences of impactful or significant experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 seldom

-0.72

respective

-0.69

igent

-0.65

 rarely

-0.63

 NEVER

-0.61

omnia

-0.60

ktop

-0.60

arcity

-0.60

perate

-0.59

lied

-0.58

POSITIVE LOGITS

âĢ

1.19

 âĢ

1.00

 somet

0.84

 something

0.83

 ðŁĳ

0.78

—"

0.76

0.75

âĺ

0.75

another

0.75

 another

0.74

Activations Density 0.917%

phrases related to observations and descriptions of things or situations

occurrences of impactful or significant experiences

No Comments

No Known Activations