INDEX

Explanations

words related to granting permission or approval

oai_token-act-pair · gpt-3.5-turbo

instances of the word "granted" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

acters

-0.62

loc

-0.61

 Dahl

-0.60

inters

-0.60

itate

-0.60

 tackle

-0.60

alph

-0.59

itating

-0.59

Pred

-0.58

 Engel

-0.58

POSITIVE LOGITS

 granted

3.71

 granting

1.93

 Granted

1.81

 awarded

1.77

 conferred

1.73

 afforded

1.71

 bestowed

1.69

 grant

1.64

 denied

1.50

 grants

1.48

Activations Density 0.012%

words related to granting permission or approval

instances of the word "granted" in various contexts

No Comments

No Known Activations