INDEX

Explanations

keywords indicating the action of 'through' or events that occur as a result of something

oai_token-act-pair · gpt-3.5-turbo

phrases that begin with "Through" or indicate causation or condition

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

 âĢ¦"

-0.67

imm

-0.61

umbn

-0.58

 tresp

-0.58

unders

-0.57

 Berks

-0.57

 harbour

-0.57

 favour

-0.57

andering

-0.56

 kindred

-0.56

POSITIVE LOGITS

itialized

0.88

cknowled

0.86

ward

0.73

ertain

0.72

 Racer

0.68

 Press

0.67

ventory

0.66

forming

0.66

 Issue

0.66

chwitz

0.65

Activations Density 0.127%

keywords indicating the action of 'through' or events that occur as a result of something

phrases that begin with "Through" or indicate causation or condition

No Comments

No Known Activations