INDEX

Explanations

adjectives followed by nouns denoting skills or qualities

oai_token-act-pair · gpt-3.5-turbo

expressions related to job performance or quality

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fears

-0.77

lines

-0.65

 disputes

-0.65

 concerns

-0.64

objects

-0.63

è¦

-0.62

 complains

-0.61

 grievances

-0.61

 boards

-0.61

 peers

-0.60

POSITIVE LOGITS

lot

1.00

job

0.98

 darn

0.92

 deed

0.88

 thing

0.85

 lousy

0.85

 damned

0.84

 good

0.83

 commend

0.81

 brisk

0.81

Activations Density 0.037%

adjectives followed by nouns denoting skills or qualities

expressions related to job performance or quality

No Comments

No Known Activations