INDEX

Explanations

references to specific quantities or levels

oai_token-act-pair · gpt-3.5-turbo

phrases that consistently include the article "a" along with other references to specific contexts or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

onis

-0.73

APTER

-0.70

apters

-0.69

eur

-0.67

 Aires

-0.67

ORTS

-0.66

imaru

-0.64

tt

-0.64

osion

-0.63

 Ending

-0.61

POSITIVE LOGITS

 glance

1.25

 distance

1.00

 rate

0.90

 snail

0.88

 disadvantage

0.86

 time

0.86

 cost

0.84

 minimum

0.82

 discounted

0.81

 fraction

0.81

Activations Density 0.050%

references to specific quantities or levels

phrases that consistently include the article "a" along with other references to specific contexts or situations

No Comments

No Known Activations