INDEX

Explanations

words related to rewards or additional benefits

oai_token-act-pair · gpt-3.5-turbo

mentions of bonuses or rewards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

URRENT

-0.80

brew

-0.76

ells

-0.76

turn

-0.74

Flickr

-0.74

URES

-0.74

avery

-0.72

 Wiley

-0.72

alter

-0.72

INK

-0.71

POSITIVE LOGITS

 bonuses

1.26

 bonus

1.26

 depreciation

0.97

 accrued

0.91

 redundancy

0.89

 athlet

0.87

 incentive

0.86

 accol

0.86

 spons

0.82

 payout

0.81

Activations Density 0.012%

words related to rewards or additional benefits

mentions of bonuses or rewards

No Comments

No Known Activations