INDEX

Explanations

words related to legal actions and consequences

oai_token-act-pair · gpt-3.5-turbo

occurrences of financial or performance-related metrics and outcomes

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

population

-0.65

estro

-0.62

resource

-0.62

 guiIcon

-0.60

 neighbour

-0.59

ãĤ´

-0.59

 ruler

-0.59

always

-0.59

othe

-0.59

cest

-0.58

POSITIVE LOGITS

 respectively

1.71

 respective

1.47

 careers

1.42

 apiece

1.31

 themselves

1.31

 roles

1.19

selves

1.11

 departures

1.07

 their

1.02

 suspensions

0.97

Activations Density 0.535%

words related to legal actions and consequences

occurrences of financial or performance-related metrics and outcomes

No Comments

No Known Activations