INDEX

Explanations

terms related to employment termination

oai_token-act-pair · gpt-3.5-turbo

occurrences of "fired" and its variations related to termination from jobs or positions

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

aque

-0.69

Yen

-0.68

lihood

-0.65

english

-0.64

 Dreams

-0.63

ournals

-0.63

 biod

-0.62

 Towns

-0.61

 Trance

-0.60

Ware

-0.60

POSITIVE LOGITS

urate

0.83

thur

0.82

balls

0.82

locks

0.76

 bullets

0.76

 firing

0.74

bats

0.74

lessly

0.74

baugh

0.74

uracy

0.74

Activations Density 0.028%

terms related to employment termination

occurrences of "fired" and its variations related to termination from jobs or positions

No Comments

No Known Activations