INDEX

Explanations

dates and time references preceded by "from"

oai_token-act-pair · gpt-3.5-turbo

the word "from" 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

faced

-0.83

ridges

-0.76

mask

-0.73

aqu

-0.73

certain

-0.72

sat

-0.70

answered

-0.70

Sys

-0.68

quote

-0.67

clear

-0.66

POSITIVE LOGITS

 afar

1.43

 whence

1.08

 abroad

1.00

 thence

0.86

 scratch

0.85

 inception

0.81

 infancy

0.78

0.76

 inside

0.75

 anywhere

0.73

Activations Density 0.199%

dates and time references preceded by "from"

the word "from" in various contexts

No Comments

No Known Activations