INDEX

Explanations

adjectives describing a negative or intense emotional state

oai_token-act-pair · gpt-3.5-turbo

adjectives that describe negative or problematic characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

helm

-0.80

ecause

-0.79

Ì¶

-0.79

uph

-0.78

uther

-0.76

cellence

-0.76

ievers

-0.74

=-=-=-=-

-0.73

--------------------

-0.72

=(

-0.71

POSITIVE LOGITS

 counterparts

1.12

 inability

1.12

 predecessors

1.11

 cousin

1.10

 reputation

1.10

 predecessor

1.09

 penchant

1.09

 foray

1.08

 creations

1.08

 namesake

1.07

Activations Density 0.270%

adjectives describing a negative or intense emotional state

adjectives that describe negative or problematic characteristics

No Comments

No Known Activations