INDEX

Explanations

precise references to the future, particularly focusing on the phrase "one day"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "one day" indicating future hopes or expectations

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

bourg

-0.66

 foss

-0.62

 Rhythm

-0.61

ãĥ¤

-0.61

umbn

-0.61

enza

-0.60

aeus

-0.60

bis

-0.59

esc

-0.59

ãĥĨ

-0.58

POSITIVE LOGITS

ago

0.91

theless

0.78

dream

0.77

elfth

0.74

Ago

0.73

 âĪ

0.67

glass

0.65

 regret

0.64

 anniversary

0.63

iatus

0.62

Activations Density 0.026%

precise references to the future, particularly focusing on the phrase "one day"

occurrences of the phrase "one day" indicating future hopes or expectations

No Comments

No Known Activations