INDEX

Explanations

words related to being selective, such as "selective," "selectively," and "selectivity."

oai_token-act-pair · gpt-3.5-turbo

terms related to selectivity and discrimination

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

loo

-0.77

ijk

-0.76

ASC

-0.76

chief

-0.75

lehem

-0.73

Jenn

-0.73

former

-0.73

Downloadha

-0.72

ruck

-0.72

FORE

-0.72

POSITIVE LOGITS

 breeding

0.89

 impulse

0.86

 serotonin

0.86

 pressures

0.85

 selective

0.82

 choice

0.80

 inhibitors

0.79

 advantage

0.75

 antip

0.74

 selectively

0.73

Activations Density 0.026%

words related to being selective, such as "selective," "selectively," and "selectivity."

terms related to selectivity and discrimination

No Comments

No Known Activations