INDEX

Explanations

terms related to fractures or breaking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

ufact

-0.70

kees

-0.65

mington

-0.65

PDATE

-0.64

VP

-0.63

racuse

-0.60

steen

-0.60

 slideshow

-0.60

viron

-0.59

heit

-0.58

POSITIVE LOGITS

ract

1.34

ors

1.06

ory

0.91

iced

0.84

ic

0.84

aneous

0.82

ual

0.82

inct

0.82

ect

0.79

ional

0.79

Activations Density 0.005%

terms related to fractures or breaking

No Comments

No Known Activations