INDEX

Explanations

sentences related to making choices or decisions

oai_token-act-pair · gpt-3.5-turbo

negative sentiments or qualifiers

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 mediation

-0.84

 resumed

-0.82

 inspections

-0.80

 reopened

-0.77

displayText

-0.76

 warnings

-0.75

 filings

-0.74

 refunds

-0.72

 reopen

-0.72

 repairs

-0.72

POSITIVE LOGITS

guy

1.18

 whoever

1.07

Whoever

1.05

 winner

1.01

someone

1.01

 somebody

0.98

 Ronnie

0.96

 Swordsman

0.95

Winner

0.95

 nominee

0.94

Activations Density 0.751%

sentences related to making choices or decisions

negative sentiments or qualifiers

No Comments

No Known Activations